Result Details

Explainable Anomaly Detection in Network Traffic Using LLM

JEŘÁBEK, K.; KOUMAR, J.; SETINSKÝ, J.; PEŠEK, J. Explainable Anomaly Detection in Network Traffic Using LLM. In 38th IEEE/IFIP Network Operations and Management Symposium, NOMS 2025. Honolulu: IEEE Communications Society, 2025. 6 p. ISBN: 979-8-3315-3164-5.
Type
conference paper
Language
English
Authors
Jeřábek Kamil, Ing., Ph.D., DIFS (FIT)
Koumar Josef
Setinský Jiří, Ing., DCSY (FIT)
Pesek Jaroslav
Abstract

Network anomaly detection is essential for modern cybersecurity, yet existing systems often generate numerous alerts without clear explanations, leading to inefficiencies and high false-positive rates. This paper proposes a novel approach that integrates Large Language Models (LLMs) with an anomaly detection framework to enhance explainability in network traffic analysis. Instead of directly detecting anomalies, the LLM only interprets already flagged anomaly events, providing insights into their potential root causes. Our method reduces LLM overusage while improving decision-making for security analysts. We evaluated our approach using real-world network traffic data, demonstrating its ability to enhance situational awareness, reduce false positives, and support more effective cybersecurity practices.

Keywords

anomaly detection, network security, network traffic monitoring, time series, large language models, explainable security

Published
2025
Pages
6
Proceedings
38th IEEE/IFIP Network Operations and Management Symposium, NOMS 2025
Conference
IEEE/IFIP Network Operations and Management Symposium 2025
ISBN
979-8-3315-3164-5
Publisher
IEEE Communications Society
Place
Honolulu
DOI
UT WoS
001556086900003
BibTeX
@inproceedings{BUT196524,
  author="Kamil {Jeřábek} and Josef {Koumar} and Jiří {Setinský} and  {}",
  title="Explainable Anomaly Detection in Network Traffic Using LLM",
  booktitle="38th IEEE/IFIP Network Operations and Management Symposium, NOMS 2025",
  year="2025",
  pages="6",
  publisher="IEEE Communications Society",
  address="Honolulu",
  doi="10.1109/NOMS57970.2025.11073574",
  isbn="979-8-3315-3164-5"
}
Projects
Flow-based Encrypted Traffic Analysis, MV, Strategická podpora rozvoje bezpečnostního výzkumu ČR 2019–2025 (IMPAKT 1) PODPROGRAMU 1 SPOLEČNÉ VÝZKUMNÉ PROJEKTY (BV IMP1/2VS), VJ02010024, start: 2022-01-01, end: 2025-06-30, completed
Departments
Back to top